MIND-IoT: Machine Intelligence and Data-Mining for IoT Threats
Kwabena Aboagye-Otchere, Jorge Castillo · 2024
The use of electric motors drive numerous industries via robust mechanical and electromechanical systems. However, their ubiquitous presence across the industry makes them potential targets for cyber-attackers, raising serious privacy concerns. Prior studies have demonstrated the feasibility of accurately fingerprinting various devices using electromagnetic signals. Yet, the specialized equipment required for those methods remain complex and expensive. This study presents MIND-IoT, a novel approach that captures and analyzes the unintended magnetic emissions surrounding electric motors. Our method achieves precise fingerprinting of motors, and determines their operational parameters by leveraging low-cost IoT devices. By focusing solely on magnetic fields, MIND-IoT offers a cost-effective fingerprinting solution while highlighting the urgent need to address associated privacy risks.